METHODS AND ALGORITHMS FOR PARALLEL CALCULATIONS USING VIRTUALIZATION TECHNOLOGIES IN MATERIALS SCIENCE PROBLEMS

The paper discusses methods and algorithms for the provision of high-performance computing resources in multicomputer systems in a shared mode for fundamental and applied research in the field of materials science. Approaches are proposed for the application of applied integrated software environments (frameworks) designed to solve material science problems using virtualization and parallel computing technologies.

Mathematical Modeling of the Dynamics of Linear Electrical Systems with Parallel Calculations

The dynamics of power systems is often analyzed using real-time simulators. The basic requirements of these simulators are the speed of obtaining the results and their accuracy. Known algorithms (backward Euler or trapezoidal rule) used in real-time simulations force the integration time step to be reduced to obtain the appropriate accuracy, which extends the time of obtaining the results. The acceleration of obtaining the results is achieved by using parallel calculations. The paper presents an algorithm for mathematical modeling of the dynamics of linear electrical systems, which works stably with a relatively large integration time step and with accuracy much better than other algorithms widely described in the literature. The algorithm takes into account the possibility of using parallel calculations. The proposed algorithm combines the advantages of known methods used in the analysis of electrical circuits, such as nodal analysis, multi-terminal electrical component theory, and transient states analysis methods. However, the main advantage over other algorithms is the use of the method based on average voltages in the integration step (AVIS method). The attention was focused on the presentation of the scientifically acceptable general principle offered to mathematical modeling of dynamics of linear electrical systems with parallel computations. However, the evidence of its effective application in the analysis of the dynamics of electric power and electromechanical systems was indicated in the works carried out by the team of authors from the Institute of Electrical Engineering UTP University of Science and Technology in Bydgoszcz (Poland).

The Influence of Dissimilarity of Comparables on the Correctness of Estimation in the Comparative Approach Procedure

The subject of this article is the still relevant issue of the adequate choice of comparables in evaluation procedures within the sales comparison approach. The text features the results of an experiment that makes it possible to validate the dependence of the accuracy of the final estimation of the price of valued objects in relation to the criteria of significance of similarity as a determinant of accepting individual objects as sufficiently similar, and the dissimilarity coefficients describing the created reference sets. Based on the example of simulated data, it has been shown that there is a clear relationship between the accuracy of the estimates made and the selection of comparables. In a set of all possible calculation variations, estimations based on selected data are characterized by more accurate calculations relative to parallel calculations carried out on the widest possible sets of comparables.

Multiprocessor modeling technologies for the applied statistical tasks

The work considers the multiprocessors technologies of modeling for Monte Carlo tasks. It is shown that only application of the modern super productive systems permitted the new way to realize the mechanism of corresponding partitioned computations. The calculating schemes that supply to provide the increase of productivity and calculations' speed effectiveness are shown. In this article the modified algorithm of parallel calculations is offered based on the Monte Carlo method. Here every calculator has its own random generator of numbers. Thus intermediate calculations come true independently on the different, separately taken blades of cluster , "calculators". The results are already processed on some separately taken master -blades ( "analyzer"). This allows to get rid from the necessary presence of router-communicator between the random generator of numbers and "calculator". Obviously, that such decision allows to accelerate the process of calculations. It is shown that the parallel algorithms of the Monte Carlo method are stable to any input data and have the maximal parallel form and, thus, minimal possible time of realization using the parallel computing devices. If it is possible to appoint one processor to one knot of calculation. Thus the realization of calculations becomes possible in all knots of net area in parallel and simultaneously.